Hot spot microvessel density and the mitotic activity index are strong additional prognostic indicators in invasive breast cancer

Aims: Recent studies have drawn attention to intratumoral microvessel densi ty (MVD) as a prognostic factor in invasive breast cancer. Various methods have been applied to assess MVD and the prognostic value of MVD in differen t studies varies considerably. Counting of microvessels in the most highly vascularized area (hot spot) of a tumour is the method most widely used. In this study we compared three counting methods. Methods and results: To assess MVD in 112 cases of invasive breast cancer w ith long-term follow-up we performed microvessel counting in the hot spot o f the tumour in four and 10 fields of vision (HS-MVD4 and HS-MVD10) and mic rovessel counting in 10 fields of vision distributed systematically over th e whole tumour area (global MVD). The HS-MVD4, HS-MVD10 and global MVD show ed good correlations with each other. HS-MVD4 provided the highest number o f microvessels (median value 71) followed by HS-MVD10 and global MVD, with median values of 58 and 39, respectively. HS-MVD4 showed the best prognosti c value for overall survival (P = 0.0001) whereas HS-MVD10 showed less (P = 0.01) and the global MVD showed no (P = 0.75) prognostic value. In univari ate analysis, the HS-MVD4 was the second strongest prognostic factor after tumour size. In multivariate survival analysis, the HS-MVD4, mitotic activi ty index (MAI), lymph node status and tumour size were found to be independ ent prognostic factors. When combining MVD4 and MAI in lymph node negative patients, none of the patients with low MVD (< 71/mm(2)) and a low MAI (< 1 0 per 10 HPF) died, in contrast to patients with a high MVD or high MAI who have a 10-year survival of 57%. Conclusions: These data suggest that the hot spot MVD in four fields of vis ion is a major independent prognostic factor for overall survival in invasi ve breast cancer. For the first time, it is shown that hot spot MVD provide s additional prognostic information to well established factors like lymph node status and the MAI, and may therefore be useful for designing treatmen t strategies in invasive breast cancer.